The effects of pH on phosphorus utilisation by chickpea (Cicer arietinum)

Chickpea (Cicer arietinum) is known to secrete organic anions. We investigated its effectiveness in obtaining P over a range of pH values. Methods We grew two cultivars of chickpea, supplied with either ammonium or nitrate nitrogen, at 10 levels of applied P, and at four initial pH values. We measured plant yields, P concentration in the tops, and rhizosphere pH. We compared the results with those previously obtained for lucerne (Medicago sativa), mustard (Brassica campestris) and rice (Oryza sativa). Results


Introduction
The roots of many plants are known to secrete organic acids and there is strong evidence that this improves their access to soil phosphate.Often secretion is mediated via cluster roots.These are ephemeral, closely-spaced groups of short roots.They secrete a burst of organic anions, especially citrate, and then die off (Shane and Lambers 2005).There are also plants that are known to secrete organic anions but without forming cluster roots.Chickpea (Cicer arietinum) is a frequently investigated example.Veneklaass et al. (2003) and Wouterlood et al. (2004a,b) found that it secreted mainly malonate, about 20% citrate, and a small amount of malate.The results of Ohwaki and Hirata (1992) are slightly different; they found citrate was the largest fraction followed by malate and malonate.There is little information on the effects of such secretion on soil pH and on the uptake of phosphate.
We recently showed (Barrow et al. 2020) that for lucerne (Medicago sativa), mustard (Brassica campestris) and rice (Oryza sativa), the optimum initial pH CaCL2 for utilisation of phosphate was just above 5.Above this value, phosphate supply was restricted by decreasing rate of uptake by roots; below this value toxicity of aluminium toxicity became important.We decided to investigate whether chickpea differed.We thought that if it acidi ed the rhizosphere, the optimum initial pH would be higher.Further, if organic acids complexed aluminium, chickpea might be better able to tolerate low pH.In the main experiment reported here we measured phosphate response curves for two cultivars of chickpea, supplying them with nitrogen either as nitrate or as ammonium, and at four different initial pH values.We measured plant yield, plant P concentration, and rhizosphere pH.
It is widely reported that plants respond to P de ciency by acidifying the rhizosphere.For example, "Under low P, plant roots release organic acids to acidify the rhizosphere" (Lei et al. 2015).Indeed, there are many examples of decreased rhizosphere pH for plants growing in soils or solutions of medium or high pH.One example is for lupin and maize growing in a calcareous loess subsoil (Ma et al. 2021).Another is for Faba beans growing in two calcareous soils (Baccari & Krouma 2023).For plants growing in a solid medium, decreases in pH occur immediately behind the root tip.Moorby et al. (1988) showed that when rape plants (Brassica napus) were deprived of P, the pH behind the root tip declined by up to 0.7 units below the initial value of 6.2.Similarly, Ho and et al. (1989) found that for roots of rape plants growing in a culture at pH 5.8 acidi cation was limited to a zone of about 1.5 cm just behind the root tip.For the rest of the solution the pH increased.In solution culture, Neumann and Römheld (1999) found that for tomato and chickpea, but not wheat, the pH of the solution dropped sharply from an initial value of 7 upon onset of P de ciency.However, it would be counter-productive to lower the pH further if it were already low.This is what Youssef and Chino (1989) found.They measured changes in the pH around roots of barley (Hordeum vulgare).When the initial pH was either above 8 or about 7, the rhizosphere pH decreased, but when the initial pH was about 5, the rhizosphere pH increased.
We similarly found that at high pH the rhizosphere pH decrease, but at low pH it increased.Our results also differ in an important aspect from those previously reported.The largest changes in pH were not associated with P de ciency.In contrast, the greater the P supply the greater the effect.At low initial pH, the rhizosphere pH increased.Again, the greater the P supply, the greater the effect.
Next, we supply some background that may help interpret this observation.
Consider the mechanism by which plants take up phosphate.Not only is this against a very steep concentration gradient, but it is also through a negatively charged membrane.There is strong evidence that uptake is accomplished by co-transporting phosphate ions and hydrogen ions (Sakano 1990, Rausch andBucher 2002).Some of the hydrogen ions are then expelled by a proton pump.The expulsion does not necessarily balance the in ow.Sakano (1990) found that phosphate uptake caused the pH of the medium to increase from pH 3.5 to values as high as 5.8, with corresponding decreases in the pH of the cytoplasm.Further, the proton balance depended on the loading; as the loading increased the ratio of protons expelled decreased; that is there was a feedback mechanism.Ulrich and Novacky (1990) also observed pH increases in the cytoplasm on uptake of phosphate.These effects would explain our observed increase in rhizosphere pH when the initial pH is low.To explain the decrease in rhizosphere pH when the initial pH is high, we suggest that chickpea has evolved a mechanism by which the activity of the proton pump depends on the external pH.

Methods
In the main experiment reported here, we used two cultivars of chickpea.One is Anuradha.This variety was previously recommended by the West Bengal government and was widely used for a long time.It is now being replaced by improved lines, one of which is AGBL 184 (Sinha et al. 2018).Both cultivars belong to the Desi group, with compound leaves and small seeds.The weight per 100 seeds of Anuradha was 9.4 g and that for AGBL 184 was 11.1 g.The P content of seeds was 4.99 mg g − 1 for Anuradha and 6.04 mg g − 1 for AGBL 184.

The soil used
We used a soil from the Regional Research Station of Bidhan Chandra Krishi Viswavidyalaya at Jhargram in West Bengal, India (22°26'58.99"N,86°59'49.23"E).This soil was previously used by Barrow et al. (2020).As indicated there, the soil was collected from well-drained site which was not used for cultivation.The average annual rainfall is 1400 mm, 81% of which falls in the monsoon months from June to early October.We collected bulk soil from the upper 20 cm during dry season; in January 2019.

Modifying the pH of soil
To raise the pH of soil to three different levels, we added powdered CaCO 3 to subsamples of soil and moistened.We incubated the samples at 60°C for 2 days to accelerate the reaction with lime (Barrow and Cox 1990).This gave us four soils with pH CaCl2 levels of 3.9, 4.8, 5.8 and 7.1.We also modi ed the pH of the bulk soil.We mixed it with powdered CaCO 3 , moistened it, and allowed it to react for several weeks before drying it ready for use.

Cultivating the plants
We grew plants in non-draining, deep, black plastic bags of 1 kg capacity.We divided each of the four soil samples of different pHs into two sets for two cultivars of chickpea used.We rst added 200 g of acidwashed white river sand.Over this, we added 500 g of bulk soil of pH CaCl2 5.2 and then 300 g of soil modi ed for pH and treated with P using KH 2 PO 4 solutions.For each pH there were ten levels of P.There were two sources of N: ammonium and nitrate.Thus, for each pH and each genotype there were 20 bags.We wetted the soil to eld capacity and sowed ten seeds of chickpea.After germination we kept ve healthy plants in all the bags.This means that the amount of P added in the seeds was (mg per pot − 1 ) Anuradha, 2.5, AGBL 184, 3.3.Seven days after germination, we added 20 mL nutrient solution containing the following basal nutrients (mg nutrient kg − 1 soil): Mn 4, Mo 1, Cu 1, B 0.2, Zn 10, Mg 6, K 100 and N 100.The amount of K added was adjusted after calculating the amount added through KH 2 PO 4 .We irrigated the plants with deionised water as required.The experiment was carried out in a net house on a roof top.
We cut the above ground parts at ground level 34 days after germination, washed them in 0.01M HCl and then in deionised water.We weighed the plants after drying in a hot air oven at 60°C to constant weight.Subsamples of 100 mg were digested using a concentrated HNO 3 -HClO 4 (v/v = 3:1) mixture for estimation of P concentrations (Murphy and Riley 1962).
We cut open the bags containing soil and roots and gently shook the plants to remove bulk soil.We regarded the soil adhering to the roots as rhizosphere soil.We cut off ten apical parts of lateral roots of 40 mm in length along with adhered soil.These were placed in pre-weighed beakers and the weights of samples were recorded.We added appropriate volumes of 0.01 M CaCl 2 solution to give a weight: volume ratio of 1:2.5 and measured the pH of this solution after allowing some time.

Estimating citrate exudation from roots
In a hydroponic experiment we used the two cultivars of chickpea plus a cultivar from a local market.Seeds were germinated and grown in quartz sand for seven days in the presence of sterilized nutrient solution whose composition was as follows (µmol): Ca(NO  2004b)..We then transferred six germinated plants of each cultivar to three large containers containing a similar nutrient solution.In these containers we grew plants for ten days.Then we transferred them to a second nutrient solution (in separate containers) containing half strength of nutrient elements for a day.Finally, we transferred the plants into nutrient free distilled water of varying pHs ranging from ≤ 4.0 to ≤ 8.0 for a day.After removing the plant roots, we measured the citric acid concentration in the solution.For any delay in estimation, 1-2 drops of chloroform were added.We measured citrate in solution using the pentabromacetone method of Dickman and Cloutier (1950).This method was designed for measuring citrate in biological materials and contained a step for precipitating protein.This was found to be unnecessary for our samples.Citrate was estimated by measuring the intensity of colour at 435-445 nm in a UV-VIS spectrophotometer (Systronics, Model: AU 2702

Describing the results
For each treatment combination, the responses were described by an equation of the form: where Y is the yield, x the P supplied, and the other symbols are parameters.The parameter a indicates the maximum to which the yields trend.The parameter c re ects the plants effectiveness in responding to P; the larger its value, the smaller the amount of P required for a given yield.The parameter d is formally the extrapolation of the response curve and may be interpreted as an estimate of the P supplied by the soil.This equation was tted using SigmaPlot 10.
When yields were related to the rhizosphere pH, the parameters a and c were replaced by functions of rhizosphere pH (Fig. 6).These equations were tted using a program written in GW BASIC.

Results
In solution culture (Fig. 1), the amount of citric acid released increased with pH.AGBL 184 released more citric acid than Anuradha.Seeds from the local market and of unknown origin released even more.
Responses to phosphate (Fig. 2) were strongly affected by the soil's initial pH.At the lowest value (pH CaCl2 , 3.9), response was markedly lower; maximum yields were not approached until at least 150 mg P − 1 had been applied.Response was also weak when the initial pH CaCl2 was 4.8.For lucerne, rice and mustard, this was close to the optimum value (Barrow et al. 2020).Strongest response occurred when the initial pH CaCl2 was 5.9; maximum yields were approached at about 40 mg P − 1 .
Figure 3 shows that in most cases there was little effect of nitrogen source on response to P. The two exceptions were for the Anuradha cultivar; at low pH, nitrate produced bigger yields, at high pH, ammonium produced bigger yields.AGBL 184 cultivar approached maximum yields that were a bit more than 20% greater than those for Anuradha cultivar.This is approximately the ratio of their seed weights and this suggests relative growth rates were similar.
Figure 4 suggests that the AGBL 184 cultivar used the applied phosphorus slightly more effectively because the values for the c coe cient were slightly higher; the larger the value of c, the smaller the amount of P required for a given yield.However, the c and d parameters are correlated; high values for one can be largely offset by low values for the other.This cultivar had low values for the d parameter, so the net effect was small.This gure also shows that the d parameter, which represents the P derived from soil and seed, formed a V-shaped relationship with pH.A similar effect has been reported for mustard, lucerne, and rice by Barrow et

al. (2020).
There were large effects of P supply on the rhizosphere pH (Fig. 5).At low initial pH the rhizosphere pH increased with P supply; at high initial pH it decreased.
For the AGBL 184 cultivar the rhizosphere pH was generally lower than that for the Anuradha cultivar (Fig. 5).The effects were largest where nitrogen was supplied as nitrate at high initial pH.This differs from the results of Veneklaass et al. (2003) who found that three commonly used Western Australian chickpea cultivars had very similar rhizosphere carboxylate concentrations.
The rhizosphere pH was generally lower when nitrogen was supplied as ammonium (Fig. 5).The effects were largest for the Anuradha cultivar at high pH.
In Fig. 6, yields are plotted against the rhizosphere pH CaCl2 .Yields were lowest at lower pH and highest near pH CaCl2 5.These responses are summarised in Fig. 4.This gure shows that effectiveness was lowest at low pH and highest at just above pH CaCl2 5.
Plots of yield against P concentration in the tissue are used to test whether growth is solely limited by P supply.When this is the case, plots fall close to a common line.Figure 7 shows that this was so for three of the initial pH values.For the initial pH CaCl2 of 3.91, plots fell slightly below the others.We think this indicates only small effect of aluminium toxicity.

Effects on pH
The effects of our treatments on the rhizosphere pH of chickpea were unusual in two respects.One is that at high initial pH, the rhizosphere pH decreased, but at low initial pH the rhizosphere pH increased.There is much emphasis on the effects of plants in decreasing the rhizosphere pH, as occurred here at high initial pH, but few observations of an increase in pH at low initial pH.To decrease the pH when it is already low would be counter-productive.
The other unusual aspect is that these effects on rhizosphere pH were closely related to the phosphate supply.The greater the supply, the greater the decrease at high pH and the greater the increase at low pH.It follows that the uptake mechanism for phosphate in chickpea is such that there is a net import of protons at low pH and a net export at high pH.The rhizosphere pH trended towards a value of about 5 from both low and high initial values.

Effects of pH
The marked decrease in pH when the initial pH was high, means that best growth occurred when the initial pH CaCl2 was 5.8, but this was associated with a lower rhizosphere pH.Best growth occurred when the rhizosphere pH CaCl2 was about 5.These effects mean that plots of the c coe cient for chickpea are almost mirror images of those for lucerne, mustard and rice (Fig. 8).For chickpea, there is a long upward slope followed by a short downward slope; for lucerne, mustard and rice, the opposite is the case.When we consider the uptake of phosphate, the differences are accentuated (Fig. 8).Low pH decreases the uptake of phosphate by chickpea much more than for the other species.This is consistent with our suggestion that the uptake mechanism for chickpea differs from that of the other species.The effects of pH would be expected to depend on the number of protons co-transported with P ions.Our results suggest that there were fewer for chickpea and therefore a greater effect of the negative membrane potential on uptake.

Organic acids
There have been several comparisons of sorption of organic anions and their effectiveness in displacing phosphate from soil or soil constituents.For example, Bolan et al. (1994) reported that P sorption was decreased in the following order: citric > oxalic = tartaric = malic > lactic = formic = acetic.Jones and Brassington (1998) reported that sorption decreased in the following sequence: phosphate > > oxalate > citrate > malate > > acetate.Sorption decreases with increasing pH (Jones and Brassington 1998).The effects of increasing pH on ion sorption are a balance between changing proportions of the reacting ion in solution and decreasing electric potential of the reacting variable-charge surfaces.This is re ected in the work of Jones and Brassington (1998) who found that at low pH oxalate was more strongly absorbed than citrate (oxalic acid, pKa 2 3.81, citric acid pKa 2 4.76).At higher pH the sequence was reversed.
In our studies of the displacement of phosphate by citric acid, we found that the main mechanism was competition between citrate ions and phosphate ions for sorption.Citrate is an effective competitor because, like phosphate, it can form bidentate links with an oxide surface.Its effectiveness is proportional to the concentration of its divalent ions in the solution (Barrow et al. 2017).This depends partly on the dissociation characteristics of the acid and partly on the concentration of cations in the solution.Calcium ions are particularly important because they are common in soil and because calcium forms complexes with carboxylic acids especially at high pH (Barrow et al. 2017).The presence of calcium decreases the pH at which maximum concentration of divalent ions occurs (Barrow et al. 2017).Although it is often thought that citrate is effective because it is a tricarboxylic acid, the third acid group plays no direct role in the reaction (Barrow et al. 2017).Its main effect is to increase the propensity for the other acid groups to dissociate; the pKa 2 for citric acid is 4.76, that for malonic acid is 5.69.This means that to compare the effectiveness of these two acids, differences in pH range need to be considered.It also suggests that it would be advantageous for a plant to secrete two acids effective over different pH ranges, and this is what chickpea does.
How P effective is chickpea?Figure 9 presents a direct comparison between the response of chickpea to P and that for lucerne, rice and mustard at similar pH values.Chickpea is indeed effective.At high pH it is almost 20 times as effective; it needs about 1/20 as much P for a similar fraction of the maximum yield.This re ects the considerable depression in response at high pH for lucerne, rice and mustard.The advantage decreases with increasing pH, but at low pH, even though there is a large decrease in effectiveness for chickpea, it is still about three times more effective than lucerne, rice and mustard.

Declarations
Funding No funds, grants, or other support were received during the preparation of this manuscript.where Y is the yield, x the P supplied, and the other symbols are parameters.The parameter a indicates the maximum to which the yields trend, as can be seen from the graph.The parameter c re ects the plants effectiveness in responding to P; the larger its value the smaller the amount of P required for a given yield.The parameter d is formally the extrapolation of the response curve and may be interpreted as an estimate of the P supplied by the soil and seed.The values for c and d are shown in Fig. 4).
Values of the parameters c and d.In the left-hand panels values derived from Fig. 2 and are plotted against the initial pH.In the right-hand panels they are derived from Fig. 6 and are plotted against the rhizosphere pH.
Effect of amount of P applied on the rhizosphere pH is affected by nitrogen source and cultivar.
Relationship between the rhizosphere pH and plant growth at 10 levels of applied P. The levels of P are those indicated in Fig. 2 Effects of initial pH and of nitrogen source on the relationship between plant growth and tissue P concentration for two cultivars of chickpea.
Comparing the effects of pH on the c coe cient for chickpea with those for lucerne mustard and rice.The ccoe cient is a measure of the effectiveness of the fertiliser; the larger its value, the smaller the amount of P required for a given yield.The upper plots are based on yield, lower ones on P uptake.

Figure 2 Effects
Figure 2

and 5 .
The lines tted are for a global equation for each panel.For these equations, the c and d parameters are replaced by functions of pH.For the c parameter a lognormal equation was used: c = a 1 + b 1 -0.5(ln(pH / c 1 )/d 1 ) 2 ) for the d parameter, a quadratic equation was used: d = a 2 + b 2 pH +c 2 pH 2 and Y = a (1-exp-c(x+d))) as for Fig.2.